Anti tumour necrosis-α therapy increases the number of FOXP3+ regulatory T cells in children affected by Crohn’s disease


Dr S. Quaratino, Cancer Research UK Clinical Centre, Southampton General Hospital, University of Southampton, Tremona Road, Southampton SO16 6YD, UK. Email:
Senior author: Sonia Quaratino


Crohn’s disease (CD) is a chronic inflammatory disease of the gastrointestinal tract. Its pathogenesis is not completely understood, though the prevailing model is that the intestinal flora drives a strong intestinal T helper 1 (Th1)/Th17 type immune response and inflammation in the genetically susceptible host. This leads to overly aggressive T-cell responses to normal bacteria causing tissue damage. Intestinal homeostasis and maintenance of tolerance to harmless antigens in the intestine has been shown to be maintained by CD4+ CD25+ T regulatory cells (Treg) in animal models of inflammatory bowel diseases. Here we investigated whether Infliximab, a chimeric monoclonal antibody directed against tumour necrosis factor (TNF)-α shown to be highly effective in the treatment of CD, has any effect on mucosal CD4+ CD25+ (FOXP3+) Tregs. Colonic mucosal biopsies from children with active Crohn’s disease treated in vivo with Infliximab and controls were analysed to determine FOXP3 expression by immunofluorescence and reverse transcription–polymerase chain reaction. We observed that FOXP3+ T cells were significantly reduced in mucosa of CD patients with active disease compared with controls and restored to normal following Infliximab treatment. The frequency of FOXP3+ cells and mRNA expression was significantly increased in CD mucosa from patients treated in vivo with Infliximab compared with CD patients treated with conventional therapies. In conclusion, we show that Infliximab treatment does not solely neutralize soluble TNF-α, but also affects activation and possibly expansion of mucosal regulatory T cells. We suggest that anti TNF-α immunotherapy can also restore mucosal homeostasis in Crohn’s disease.


Crohn's disease


inflammatory bowel disease




rheumatoid arthritis


tumour necrosis factor


T regulatory cells


Crohn’s disease (CD) is a chronic inflammatory bowel disease of unknown aetiology. Perturbation of the intestinal mucosa homeostasis is believed to play a pivotal role, but the pathogenesis of CD is still not fully understood. One current hypothesis suggests that the chronic inflammation in CD is a consequence of dysregulation of the adaptive immune system towards components of the intestinal flora.1 This leads, in genetically predisposed individuals, to an immunological imbalance characterized by an excessive production of pro-inflammatory cytokines. How tolerance at mucosal sites is regulated is not exactly known, but it is clear that many different cells, cytokines and other factors are involved.2

CD exhibits an immunological response characterized by an exaggerated CD4 T helper 1 (Th1)/Th17 type phenotypes. The pattern of cytokines expressed by lamina propria CD4 T cells is characterized by an increased production of interferon-γ (IFN-γ), tumour necrosis factor-α (TNF-α) and interleukin-17 (IL-17).3

TNF-α is considered the prototypical pro-inflammatory cytokine secreted by activated macrophages and monocytes and chronically activated T lymphocytes.4 Despite the existence of redundancy among mediators of inflammation, a hierarchy of importance has emerged with TNF-α as a key effector and regulatory molecule in Th1/Th17 responses.5

The mechanism by which TNF-α regulates inflammation in the intestine of a Crohn’s patient is likely to be complex and multifactorial. Anti-TNF-α treatment resulted in down-regulation of TNF-α and IFN-γ production in the mucosa of Crohn’s patients. Th1-type T cells were not eliminated but their function was reduced in inflamed mucosa to a level comparable with that seen in uninflamed mucosa.6 This finding supports the notion that TNF-α augmentation of Th1/Th17 function in the mucosa is critical for disease pathogenesis. Therefore, TNF-α might be of particular importance in the mechanism leading to inflammation in CD, as it regulates the production of other pro-inflammatory mediators; hence, the rationale for its use as a therapeutic target.

Several TNF-α neutralizing antibodies and fusion proteins have been reported to be clinically beneficial in chronic inflammatory diseases.7,8 One of these, Infliximab, a chimeric monoclonal antibody (three quarters human-constant region immunoglobulin G1 (IgG1), and one quarter mouse variable region Fv), is effective in inducing clinical remission in steroid-refractory CD, with positive effects in 60% of CD patients.9 Infliximab is able to neutralize both forms of TNF-α, by either binding the transmembrane form (mTNF-α) and/or by blocking the soluble form (sTNF).10 Other anti TNF strategies which block TNF signalling (e.g. Etarnecept) that is effective in the improvement of rheumatoid arthritis (RA), have been shown to be ineffective in Crohn’s.9 An explanation to this could be that Etarnecept binds just the sTNF and not the mTNF as Infliximab does.

Inappropriate immune responses are usually kept at bay through a mechanism of active suppression operated by regulatory T cells.11,12 Naturally occurring CD4+ CD25+ T regulatory cells (Tregs) are essential for the maintenance of mucosal tolerance and several studies have shown that CD4+ Tregs are also capable of preventing or curing colitis in animal models.13–15 CD4+ CD25+ Tregs express FOXP3, a member of the fork-winged helix family of transcription factors, which is a master gene and plays an important role in their development and function.16,17

Recent studies in RA have revealed a novel and potentially important immunoregulatory action of TNF-α suggesting that this cytokine is able to down-regulate CD4+ CD25+ Treg function.18 This results in a decrease in FOXP3 mRNA and protein expression by Tregs.

So far, there are few studies describing the role of CD4+ CD25+ Tregs in human inflammatory bowel and Crohn’s diseases. One study reported the suppressive activity of CD4+ CD25+ T cells present in the intestinal lamina propria.19 Another demonstrated that gut derived CD4+ CD25+ T cells from CD patients expressed FOXP3 and were able to suppress the proliferation and the production of the pro-inflammatory cytokines cytokine TNF-α and IFN-γ of autologous CD4+ T cells.20

Lamina propria CD4+ CD25+ Treg cells might be involved in mucosal homeostasis in inflammatory bowel disease (IBD).15 How the mucosal homeostasis is broken in CD so that there is the development of the disease is unknown. As already observed in other autoimmune diseases (such as multiple sclerosis or RA) a reduced function of Tregs might be the reason for the breakdown of immunological self-tolerance.21,22

The aim of the present study is to investigate the role of Tregs in CD and whether TNF blockade by in vivo treatment with Infliximab has any effect on the mucosal Tregs. Our results clearly show that treatment with Infliximab restores high levels of CD4+ CD25+ Tregs in the mucosa of children affected by Crohn’s disease.

Materials and methods

Patients and biopsy specimens

Biopsy specimens from pediatric patients with Crohn’s and controls were taken during colonoscopy at the Gastroenterology Unit, Great Ormond Street Hospital, London. Colon specimens from seven CD patients treated with Infliximab, from five CD patients treated with conventional therapies that involve remission induction with enteral feeds followed by remission maintenance with azathioprine and a 5-aminosalicylate preparation and from four controls (children being investigated for constipation in whom inflammation was absent in routine laboratory histology) were available for study. CD was diagnosed by established clinical and histopathological criteria. Fully informed consent was obtained from the parents of all patients. Ethical approval was granted by the Great Ormond Street Hospital REC.

Each biopsy specimen was washed in 0·15 mol/l sodium chloride and examined with a dissecting microscope. One specimen from each patient was oriented and embedded in OCT, snap frozen in isopentane cooled in liquid nitrogen, and then stored in liquid nitrogen until cryosectioning.

Immunostaining on mucosal samples

Five μm-thick cryostat sections of each intestinal mucosa sample from the colon of CD patients and controls were cut. Sections were fixed in 4% PFA then washed in Tris-buffered saline (TBS) (pH 7·4). Sections were blocked for non-specific binding with 10% goat serum and then incubated overnight at +4° with a mouse monoclonal antibody anti FOXP3 (clone 236A/E7) (Abcam, Cambridge, UK) followed by incubation with a secondary antibody goat anti-mouse biotinylated (DAKO, Ely, UK) and then by streptavidin–fluoroscein isothiocyanate (FITC; DAKO) or alternatively by streptavidin–horseradish peroxidase (HRP; DAKO) for immunohistochemistry staining. Immunostaining was visualized and quantified with a Zeiss Axioplan2 imaging microscope.

FOXP3 expression determined by reverse transcription–polymerase chain reaction (RT–PCR)

Intestinal biopsies were collected from CD patients with active disease and from CD patients who had been treated in vivo with Infliximab. Colonic biopsies were stored in RNAlater (Ambion, Austin, TX) to prevent RNA degradation. Total RNA was isolated using the Trizol reagent method (Gibco, Paisley, UK). Total RNA was extracted using an RNeasy Mini Kit (Qiagen, Hilden, Germany) according to the manufacturer’s instructions. The amount and purity of the obtained RNA was determined by measurement of optical density at 260 and 280 nm. RT–PCR was performed in a two-step procedure. The cDNA synthesis was carried out with 10 μl total RNA using reverse transcriptase and oligo/dT. The second step PCR was performed in a 50 μl reaction volume containing 2 μl cDNA, 1·5 mm MgCl2, 0·1 mm deoxynucleotide triphosphate (dNTP), 5% dimethyl sulphoxide (DMSO), 1·5 μl NH4 buffer 10×, DNase/RNase-free water, 0·1 μl TaqPol, and 1 μm of each primer. Primer sequences for Foxp3 were: Foxp3 forward: 5′-TCA AGC ACT GCC AGG CG-3′ and Foxp3 reverse: 5′-CAG GAG CCC TTG TCG GAT-3′. As control, mRNA content for the housekeeping gene glyceraldehyde-3-phosphate dehydrogenase (GADPH) was analysed using the following primers: GAPDH forward: 5′-AGC CAC ATC GCT CAG ACA C-3′ and GAPDH reverse: 5′-GAG GCA TTG CTG ATG ATC TTG-3′. After 2 min and 30 s at 92°, 40 cycles of amplification were followed by 5 min extension at 72°. Each cycle included denaturation at 92° for 30 s, annealing at 55° for 30 s and extension at 72° for 30 s. The PCR products, alongside a 100 kb ladder, were run on 2% agarose gel stained with ethidium bromide and visualized by ultraviolet illumination and the image captured using Biorad Chemidoc.

Statistical analysis

Parametric data are expressed as the mean ± SD of data obtained from at least four to five separate experiments. The statistical significance of differences between group means was determined using Student’s t-test using Prism 4 software (Graph Pad, San Diego, CA). A P-value of <0·05 was considered statistically significant difference between groups.


FOXP3+ cell numbers are reduced in the mucosa of CD patients with active disease compared with controls

To study the Tregs in CD and to try to understand the role, if any, of these cells in the pathogenesis of the disease, we firstly measured Treg frequency in CD and control mucosa by measuring the expression of FOXP3+ cells at the site of inflammation. Colonic mucosal biopsies from active CD patients treated with conventional therapies and from controls were compared (Fig. 1). We observed that the frequency of FOXP3+ cells/mm2 of lamina propria was significantly reduced in the mucosa of CD patients with active disease (n = 4, with mean values: 7·000 ± 1·472) compared with controls (= 4, with mean values: 12·25 ± 1·250; P = 0·0347*) (Fig. 1).

Figure 1.

 FOXP3+ cells are decreased in CD mucosa compared with controls. Colon mucosa from CD patients treated with conventional therapies (n = 4) and controls (n = 4) were stained for FOXP3 expression and analysed by immunohistochemistry. The quantification is based on the number of positive nuclei per area (mm2) of lamina propria. All the data reported in this figure were reproduced in at least five separate experiments. Statistical significance was analysed by Student’s t-test.

FOXP3 increases in mucosa of CD patients treated in vivo with Infliximab compared with mucosa of untreated CD patients and controls

We investigated whether anti-TNF-α therapy could influence the local frequency of CD4+ CD25+ Tregs and their FOXP3 expression. Biopsies from CD patients treated with Infliximab, or conventional therapies and controls were analysed to investigate whether there was any significant difference either in the frequency of number of FOXP3+ cells or in the expression of FOXP3 transcript. Nuclear expression of FOXP3 was higher in mucosa of CD patients treated with Infliximab (with mean values: 24·00 ± 3·071 n = 7) compared with mucosa of untreated CD patients (with mean values: 7·000 ± 1·304 n = 5) and controls (with mean values: 11·75 ± 1·377 n = 4) (Figs 2 and 3).

Figure 2.

 Expression of FOXP3 in colon mucosa lamina propria of controls, active and Infliximab treated Crohn’s patients. Colon biopsies from normal controls, Crohn’s patients with active disease or treated in vivo with Infliximab were analysed. FOXP3 staining (green) was visualized by immunofluorescence and analysed using fluorescence microscopy. Arrows show nuclear transcription factor FOXP3 expression. A representative example of FOXP3 expression. All patients were tested and similar results were obtained from the different groups. Data are representative of at least five separate slides. Original magnification ×100.

Figure 3.

 FOXP3+ nuclei expression in Lamina propria of colon mucosa from controls, active and Infliximab-treated Crohn’s patients. The quantification is based on the approximate number of positive nuclei per area of lamina propria. Results from seven CD patients treated in vivo with Infliximab, five CD patients treated with conventional therapies and four controls are shown. Statistical significance was analysed by Student’s t-test. Plots show mean values with SD bars.

In these experiments RT–PCR was also performed to quantify the FOXP3 mRNA transcript in mucosa from CD patients treated or not with Infliximab. A significant up-regulation of FOXP3 mRNA was detected in colonic mucosa from CD patients treated with Infliximab compared with mucosa from patients treated with conventional therapy (Fig. 4). These data show that during active Crohns’ disease the frequency of Tregs is significantly reduced, mirrored by a reduction of the FOXP3 mRNA, a marker of Tregs. Treatment with anti-TNF antibody (Infliximab) in vivo significantly increased the frequency of Tregs in the tissue and their expression of FOXP3 mRNA and protein. These results suggest that treatment with Infliximab leads to the induction of regulatory T cells in Crohn’s disease.

Figure 4.

 FOXP3 mRNA transcript was increased in the mucosa of Crohn’s patients treated in vivo with Infliximab. Colonic biopsy specimens from Crohn’s patients treated in vivo with Infliximab or conventional therapy were analysed. mRNA isolated from intestinal biopsies was subjected to RT–PCR and analysed for FOXP3 transcript. FOXP3 mRNA was increased in the mucosa of CD patients treated in vivo with Infliximab. FOXP3 mRNA expression was normalized to the housekeeping gene GAPDH. A representative example of Foxp3 mRNA determination in a patient treated with Infliximab and a patient treated with conventional therapy is shown. Similar results were obtained from all other patients treated with or without Infliximab. One out of four independent experiments is shown.


The potential of Treg cells to modulate immune responses has led to considerable interest in their use for the treatment of autoimmune disease and chronic inflammatory diseases. Regulatory T cells prevent biological ‘friendly fire’ by ensuring that the T cells do not attack the body’s own tissues. Failure of the regulatory T cells to control the frontline fighters leads to excessive immune responses, as in Crohn’s disease (CD). Here we have shown that in active untreated CD there is a significant reduction of the frequency of Treg at the site of inflammation and that treatment with Infliximab is able to expand the local Tregs in CD patients. Despite the small sample size, these results provide new understanding into the role of FOXP3 Tregs within the context of CD. The pro-inflammatory environment, rich with TNF-α, generated during the active status of CD, can hamper viability or expansion of Treg, reducing their frequency in the lamina propria. This would result in a reduced regulatory activity of these cells and further amplification of the inappropriate immune response. During initial tissue invasion by colonic micro-organisms, exuberant TNF-α production may limit the activity of Tregs by binding to the tumour necrosis factor receptor-2 (TNFR2), and promote induction of immune reactivity and the effector phase of lymphocyte responses. It is conceivable that TNF-α has a direct effect on regulatory T cells’ viability, such as the induction of apoptosis through the binding mTNF-α with TNFR2.23 So in CD the loss of mucosal homeostasis with the increase of T-cell proliferation and the apoptosis of Tregs might be caused by the effect of TNF-α. The increased number of FOXP3+ Tregs after TNF-α neutralization with Infliximab, may be explained by a decrease in mTNF-α that results in reduced activation of Tregs through TNFR2 and reduction of Treg apoptosis. With the enhancement of Tregs function immunological tolerance is restored.

Another explanation of the increase of FOXP3+ cells could be caused by the fact that Infliximab induces apoptosis of pathogenic T cells through the mitochondrial pathway, pathway that Tregs are not sensitive to. This might explain why treatment of patients with Infliximab does not induce apoptosis of Tregs. In this way, TNF-α may play an important instructive role in controlling adaptive immunity. The manipulation of TNF-α signalling in Tregs may result in novel therapeutic approaches to augment the limited and/or inadequate function of these FOXP3+ Tregs in CD. Treatment in vivo with anti TNF-α antibody resulted in a significant clinical benefit although it could not be proved whether such change was an epiphenomenon or a primary therapeutic event. Unfortunately, for ethical reasons we could not obtain peripheral blood from these children with Crohn’s disease for further investigations.

Further studies on larger groups of patients with CD receiving anti-TNF-α therapy are required to confirm these data, elucidate the mechanisms that underlie these observations and relate the clinical and immunological responses. Restoration of the function of Treg cells or transfer of fully competent Treg cells could be a useful therapeutic tool in the treatment of CD.


The study has been supported by CICRA and Coeliac UK.